The present invention relates to an assay for testing inhibitors of the interaction between proteins p53 and hdm2.
The protein encoded by the human double minute 2 gene, hdm2, forms a complex with the tumor suppressor gene product p53 both in vitro and in vivo. In some human cancers hdm2 is overexpressed and binds most of the cellular p53. Formation of this complex is favoring nucleoplasmic transformation since the complexed p53 looses the tumor suppressor activity. Compounds which prevent the interaction between p53 and hdm2 will release p53, promoting its tumor suppression activity within these cancerous cells. Similar results could also be obtained with animal cancer cells, e. g. in mouse. The mouse homologue to hdm2 is mdm2.
To search for inhibitors of p53 -hdm2 interaction, a high throughput primary binding assay, for example ELISA, can be used to select compounds and to initiate a medicinal chemistry program. However, assays which can be used for such a primary screening of p53-hdm2 binding have the disadvantage that artefacts may occur, i. e. wrong positive reaction can be obtained because of artefactual results due to the chemical properties of the tested substances.
In addition, compounds which inhibit the interaction between p53 and hdm2 can also alter p53 specific DNA binding, which is a totally undesired effect because DNA binding is a prerequisite of p53 tumor suppressor activity.
For example, p53 is sensible to several chemical agents that inhibit its activity. The main criteria of activity of p53 is the DNA binding which reveals that the protein is properly folded and not aggregated or unfolded. Compounds which like metal chelators precipitate p53 might be considered as true inhibitors of the p53 -hdm2 interaction in a classical binding assay because the precipitated p53 cannot form complexes.
Therefore, additional testing of the impact of the substance on p53-DNA binding is important because compounds inhibiting p53-DNA binding are not good candidates for therapeutic uses. However, in a high through put assay it is not possible to test whether a compound which inhibits p53-hdm2 interaction prevents p53 specific DNA binding or disturbs the p53 conformation so that p53 can no more fulfill the desired biological function.
To avoid these two problems and to start a chemistry program based on more relevant lead compounds, the use of a good confirming assay is crucial.
A confirming assay according to the present invention could, for example, be a gel shift assay. A gel shift of a p53-DNA complex in an agarose gel after incubation with adeno-virus E1B protein is described in Yew et al. [Genes and Dev. 8, 190-202, 1994].
In Wang et al. [PNAS 91, 2230-2234, 1994] the detection of the binding of HB virus X protein to p53 by measuring the inhibition of p53-DNA binding is described. However, none of the prior art publications describes an assay in which the DNA binding property of p53 remains if p53 is complexed with a double minute 2 protein.
So far, all the in vitro assays described in the literature to study the interaction between p53 and hdm2 are immunoprecipitation assays for testing the binding of hdm2 to p53 [Leng et al. , Oncogene 10:1275(1995)]. None of these assays simultaneously show that hdm2 binds to p53 and does not disturb its specific DNA binding.
In the present invention it was surprisingly found that p53-DNA binding is maintained after complex formation with hdm2 and that it is possible to measure in one and the same reliable assay the effect of a substance on both the p53-hdm2 and p53-DNA binding.
It is the object of the invention to provide a reliable test method for compounds which inhibit the formation of complexes between hdm2 and p53 but which do not inhibit binding of DNA to p53 or disturb the p53 conformation so that p53 can no more fulfill the desired biological function.
The present invention concerns a new assay which allows the identification to compounds which inhibit the formation of complexes between a product of the double minute 2 gene (xe2x80x9cdm2xe2x80x9d), for example human hdm2 or mouse mdm2, and p53 but not between p53 and DNA. Both the complex formation of labeled DNA, C-terminally truncated p53 and dm2 and disruption of dm2 from the labeled DNA-p53 complex by an inhibitor of the p53-dm2 interaction can be detected by a gel shift assay procedure. This assay permits the selection of compounds which, besides their inhibitory property, do not alter p53 specific DNA binding and do not disturb p53 conformation required for DNA binding or formation of active tetramer.
The invention further concerns a test kit for testing the effect of a substance on the binding of a dm2 protein to p53, comprising (a) a p53 or functional equivalent thereof having DNA-binding, oligomerisation and hdm2-binding properties, (b) a hdm2 or functional equivalent thereof having the p53 binding domain, and (c) a DNA sequence specifically binding to the p53 binding domain.
The present invention concerns a test method for a substance inhibiting the formation of a complex between p53 and a product of the double minute 2 gene (xe2x80x9cdm2xe2x80x9d), for example human (h)dm2 or mouse (m)dm2, while not inhibiting the formation of a complex between p53 and DNA. The method comprises measuring complex formation in a mixture of p53, dm2 and DNA binding to p53 in the presence and in the absence of a substance to be tested. In the presence of the desired property of the tested substance, a complex between p53 and DNA is formed (xe2x80x9cdouble complexxe2x80x9d), while in the absence of the desired property either a complex between p53, DNA and dm2 (xe2x80x9ctriple complexxe2x80x9d; if no inhibiting activity is present) or no complex (if the tested compound inhibits both the dm2-p53 and p53-DNA complex formation or if the tested substance destroys the p53 conformation so that it is no more DNA binding) is formed. While any method being able to discriminate between the different conditions (triple complex, double complex, no complex) is suitable for performing the present assay, in a preferred embodiment of the invention the assay performed is a gel shift assay.
Thus, the test system essentially comprises a p53, a dm2, and DNA. While the use of the human proteins or active variants thereof is most preferred, the invention is not limited to the use of the human proteins. The corresponding proteins from other species can also be used, e. g. from mouse. However, it is preferred that both the p53 and dm2 protein used in the assay originate from the same species.
For performing the present invention, a p53 protein must be used which both is able to bind DNA and dm2.
p53 according to the present invention can be a recombinant form of p53 or purified from the original organism. It is, however, not necessary to use a full length p53 for performing the present invention. Accordingly, the p53 form used herein also means any useful variant or fragment of p53, preferentially of human p53. The features of such a useful variant or fragment are clear from the description hereinafter.
For DNA binding, p53 must be able to form tetrameric complexes. Consequently, for DNA binding both an active p53 DNA binding domain (e. g. residues 102-292 of p53) and a p53 functional oligomerisation domain (e. g. residues 325-356 of p53) must be present in the p53 form used in the present invention. For improving DNA binding properties of p53, the protein can be activated by interaction with a specific antibody (for example the monoclonal antibody Pab421 known in the art which binds to the amino acid stretch between amino acid 372 and 380 of the human p53), phosphorylation by kinases (casein kinase 11 phosphorylating Ser392 of human p53 or protein kinase C phosphorylating Ser 370 and Ser 375 of human p53) or, more preferably, truncation of its C-terminus (deletion of maximal 38 amino acids of the C-terminus of the natural p53 sequence). An example for the latter is p53D30, i. e. natural p53 lacking the C-terminal 30 amino acids, used in the Examples hereinafter.
In another embodiment of the invention, a p53-DNA complex for studying inhibitors of dm2-p53 interaction can also be obtained by the use of high affinity binding DNA elements (like the RGC and the BC sequences described in Kern et al. , Science 252:1708, 1991, and Halazonetis et al., EMBO J 12:1021, 1993, respectively). p53 can be directly purified from various sources (for example from bacteria, baculovirus or mammalian cells).
For being able to form complexes with hdm2, a human p53 suitable for use in the present invention contains residues 1 to 52 of the natural p53 sequence, more preferentially residues: 18 to 23, even more preferentially residues 19, 22 and 23.
The concentration of p53 which is preferentially used in the present invention depends on the amount of dm2 used in the assay. Normally, a five fold excess of dm2 protein is used. A very clear signal in detection of radioactive label is obtained with about 50 to 100 ng of p53D30. However, it is also possible to use higher amounts if it is possible to tolerate in the test assay that some of the proteins precipitate.
A dm2 protein for use in the present invention can be recombinant or purified from the original organism.
A dm2 protein for use in the present invention can be either the full length form or a truncated form or any hybrid protein which contains the minimal p53 binding domain of the dm2. dm2 in context with the present invention means preferentially a dm2 from the same species as the p53 used in the assay is derived from. In particular, if human p53 is used, a human dm2 (hdm2) or analogue thereof containing the minimal p53 binding domain is used, and if mouse p53 is used, a mouse dm2 (mdm2) or analogue thereof containing the minimal p53 binding domain is used. For hdm2, the region from residue 1 to residue 102 of the natural sequence is identified so far as minimal p53 binding domain. For example, in an embodiment of the invention a fusion protein consisting of the N-terminal 188 amino acids of hdm2 (comprising the p53 binding domain) and the full length glutathione S-transferase from Schistosoma japonicumprepared in the Examples (named herein G-M fusion protein) can be used for the assay. The fusion protein is obtainable by expressing the DNA encoding the N-terminal 188 amino acids of hdm2 in the expression vector pGEX-2T (Pharmacia).
The DNA element of the test system can be any DNA fragment which specifically binds to p53, e. g. such containing a p53 binding element degenerated or not. It can be a synthetic oligonucleotide, a DNA fragment isolated from living organisms or a DNA element inserted in a plasmid.
The optimal p53 to hdm2 ratio may vary depending on the purity and specific binding activity of the used proteins and, thus, should be determined for each protein variant used.
In the case that a gel shift assay is performed, the DNA element should be such that a DNA band can be detected which shifts in the gel when the DNA is incubated with p53. For detection, the DNA element can be either radiolabeled or possibly labeled by a non radioactive method.
For obtaining a satisfactory detection signal, p53 should be satured with DNA. For example, the KD of full length p53 activated by antibody Pab421 is about 5xc3x9710xe2x88x9210 M
For 50 ng p53D30 used in the Examples, 0.1 to 0.5 pmole of DNA should be sufficient.
For gel shift assay, the gel can be an agarose gel or, preferably, a native polyacrylamide gel, preferably such having 4 to 5% acrylamide. The buffer can be any buffer in which p53 is active for specific DNA binding since the complex p53 -DNA is less stable than the p53 -hdm2 complex. Preferred buffers are HEPES in a concentration of 20 to 50 mM or buffered Tris solution in a concentration of 10to 50 mM.
In a preferred embodiment the pH of the buffer is 7.1 to 8.0. In a preferred embodiment of the invention a salt is present in the buffer. If a salt is used, it should preferentially be KCl at 50 to 100 mM or NaCl at 50 to 175 mM. Moreover, the buffer can optionally contain a substance selected from Glycerol (up to 20%), DTT (up to 0.5 mM), MgCl2 (e. g. about 6 mM), ZnSO4 (e. g. about 0.1. mM), ZnOAc (e. g. about 0.1 mM), detergent NP40 (up to 0.1%), Triton X-100 (up to 0.1%), bovine serum albumin (up to 1 mg/ml), EDTA (up to 1 nM), and a competitor DNA, e. g. poly dl-dc, poly dA-dT or salmon sperm DNA, e. g. in a concentration of 25 to 100 xcexcg/ml.
The invention further concerns a test kit for testing the effect of a substance on the binding of a dm2 protein to p53, said test kit essentially comprising (a) a p53 or functional equivalent thereof having DNA-binding, oligomerisation and hdm2-binding properties, (b) a hdm2 or functional equivalent thereof having the p53 binding domain, and (c) a DNA sequence specifically binding to the p53 binding domain. The preferred ingredients of the test kit are as above. The test kit optionally contains instructions for its use.